Gallup surveyed more than 100 chief human resources officers who sit on its Global CHRO Roundtable and found something worth sitting with. Nearly all of them, 99%, say artificial intelligence is important to their company's strategy. Yet half of those same executives admit they are not confident their managers can actually guide employees through the change.
That split matters because managers are the ones who translate a company strategy into daily instructions. If the people running the strategy do not trust the people executing it, the technology rollout stalls no matter how much money went into buying it.
This is not a one off finding. A Gartner survey of HR leaders found that only 8% believe their managers currently have the skills to use AI effectively, even as a third of those same organizations expect employees to perform better simply because AI is now available to them. ManpowerGroup found an even starker number: only 3% of organizations say their leaders are highly prepared to manage AI powered teams, while 78% of organizations report employees are worried about what AI means for their jobs.
Managers are not hiding from this either. In a joint survey by Indeed and YouGov, 43% of managers said they feel poorly equipped or not equipped at all to lead a workforce that companies now expect to be fluent in AI tools. Over half of workers in that same survey said they are not getting the training they need.
The gap creates a real business problem, not just an awkward statistic. Mercer surveyed more than 8,000 employees globally and found that fewer than 20% had heard anything from their direct manager about how AI would affect their job or the business. When managers stay quiet, employees fill that silence with worry about being replaced, which slows down adoption far more than any technical limitation does.
There is a second, less obvious layer here. Gallup's own tracking shows manager engagement, meaning how invested managers feel in their own jobs, dropped from 31% in 2022 to 22% in 2025. A manager who is disengaged from their own role is unlikely to become an enthusiastic AI champion just because HR asked them to. Training a manager on a new tool does not fix a manager who is already stretched too thin.
CHROs are trying to respond. Most of the surveyed executives say they are encouraging peer learning and experimentation, and more than half are now offering AI training aimed specifically at managers. Fewer have gone further to build AI skills into job expectations or performance reviews, which suggests the effort is still at the encouragement stage rather than becoming a real requirement.
The bigger picture backs up why this caution is warranted. A widely cited MIT study on generative AI in business found that only 5% of organizations report measurable financial return from their AI spending so far. That number lines up neatly with the management gap: buying AI tools was never the hard part. Getting a tired, stretched manager to explain the tool to a nervous employee is the hard part, and no software update fixes that on its own.
For any business watching this from the outside, the lesson is not to rush out and buy an AI certification course for middle managers. It is to check whether those managers have the time, information, and basic confidence to have an honest conversation with their teams about what is changing and why. That conversation, done badly or not at all, is turning out to be the actual bottleneck.